Topic: Sinking in sunk costs

PM Clinic: Week 18 discussion summary

Compiled: 3/7/2005

The Situation:

My team was faced with the choice of buying expensive off-the-shelf software or writing it in-house - we decided on the latter course since it was an important tool (a software performance analyzer), was within the range of our expertise, and is something we expect to customize in the future. After five months of development (a month beyond the initial schedule), the product is still not working properly, and is significantly far away from completion (another 8 weeks by current estimates). The in-house development costs have already exceeded the costs of the off-the-shelf product. When should the PM face reality and persuade management to buy the product? Or should we throw good money after bad and wrestle the in-house product to completion?

- Sinking in sunk costs, Vancouver BC


  • See the situation fixing a train wreck in progress.
  • You still need to flesh out the plan for doing the work yourself. Only then can you evaluate this approach vs. purchasing the off the shelf product.
  • Kremer strongly recommended spending lots of time with the off the shelf program. Once you've committed to it, it will be expensive to undo your choice. If you can, beta test using the software in the real work environment and seriously evaluate it before you commit to it.
  • Berkun pointed out that all the MBA advice says to ignore sunk cost. You never throw good money after bad. Make each decision on its own merits, not on what it will possibly recover from past mistakes (or put another way, trying to undo mistakes is not the same thing as progress).
  • Make vs. Buy decisions should hinge on how central the software is to your organization. If it's a core part of the business, you want to make and own. If it's a small part of the business, you want to rent, borrow or outsource. (E.g. A good italian restaurant should not buy pre-made tomato sauce, but might buy prepackaged silverware).


J Kremer, Scott Berkun (editor)

All content copyright 2005. Scott Berkun. RSS Feed